CN102131038A - Image compressing apparatus, compressed image outputting apparatus and image compression method - Google Patents
Image compressing apparatus, compressed image outputting apparatus and image compression method Download PDFInfo
- Publication number
- CN102131038A CN102131038A CN2011100253016A CN201110025301A CN102131038A CN 102131038 A CN102131038 A CN 102131038A CN 2011100253016 A CN2011100253016 A CN 2011100253016A CN 201110025301 A CN201110025301 A CN 201110025301A CN 102131038 A CN102131038 A CN 102131038A
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- concentration
- color
- black
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/41—Bandwidth or redundancy reduction
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Of Band Width Or Redundancy In Fax (AREA)
- Facsimile Image Signal Circuits (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention provides a color image processing apparatus. In the color image processing apparatus, in a compression process section (an image compressing apparatus), when an original document is compressed, the edges of black texts and/or line drawings in the original document is detected, then a foreground layer (a first image) is generated in which the detected edges are expressed in the form of a binary image, and then lossless compression is performed on the foreground layer. Further, a background layer (a second image) is generated in which a density difference of the edges from the other parts in the original document is reduced, and then lossy compression is performed on the background layer. In the image obtained by decompressing the compressed file, the contours of black texts or line drawings are expressed clearly, and at the same time, noise caused in association with compression is reduced when performing the lossless compression by shortening the density difference.
Description
Technical field
The present invention relates to the method for compressing image of compressing image data, more particularly, relate to by being literal and background separation of images, compress with suitable method respectively, expeditiously the image compressing device of compressing image data, compressed image output device, method for compressing image, computer program and recording medium.
Background technology
The letter image of reading and recording on paper etc. optically, the image processing techniques that generates the view data of electronic data are used in scanning means, picture unit, copy device widely or are had both in the field of compounding machine of these functions.The view data that is generated is with transmitting-receivings such as facsimile or Emails, and it is medium perhaps to be stored in database, is utilized in the various uses.The view data that reading images generated is general because picture size is big optically, therefore in order to receive and dispatch effectively or to store, need compress view data.
As being used to one of compress technique that realizes high compression rate, having MRC (mix raster content Mixed Raster Content) and wait compress technique based on layer separation.The compress technique of separating based on layer is the foreground layer that shows the literal lines from image formation sheet, according to the foreground layer that is generated, is separation of images foreground layer and background layer, by the compression method that uses each self application foreground layer and background layer is compressed, and finally generates the technology of high compression image.Here, so-called foreground layer, be the layer of the prospect of literal in the presentation video and/or lines, generally use JBIG (the associating bianry image Joint Bi-level Image Experts Group of expert group), MMR (improved two dimensional compaction coding Modified Modified Read) or LZW reversible compress techniques such as (Lempel Ziv Welch) to compress.
On the other hand, background layer is the layer of the background of the picture material beyond expression literal and/or the lines, generally uses JPEG non-reversible compress techniques such as (joint photographic experts group) to compress.In the technology of in patent documentation 1 (spy opens the 2003-18413 communique), putting down in writing, to text filed each that comprises in the colored document image, the color that extraction is used in it is text filed, generation comprises the text filed image that comprises that has used with a kind of color, carry out the MMR compression for comprising image, this is comprised retinue (Pallet) information of image and comprises image information add as title.In addition, gazing under the text filed situation that a kind of color only arranged, this is being gazed at the text filed MMR of carrying out compression,, carrying out reversible compression gazing under the text filed situation that multiple color arranged.Palette information and text message are added as title in compression result.And then, in patent documentation 1, put down in writing and used text filed in the specific colored document image of text message, generate with the color of regulation and smear specific text filed image (being equivalent to background layer), the image that is generated is carried out the technology that JPEG compresses.
Compression based on non-reversible compress technique is compared with the compression based on reversible compress technique, and the image quality of the image after the compression is easy to degenerate.But therefore non-reversible compress technique according to the purposes of compressed image, can improve compression ratio because the control of compression ratio is simpler when preferentially dwindling data size, perhaps reduction compression ratio when preferential image quality high-quality.Otherwise therefore reversible compress technique is difficult to improve compression ratio owing to be difficult to control compression ratio.
Summary of the invention
When literal in separate picture and background, generally be to extract whole literal shape.In order to extract whole literal shape, can by with between the rising edge at edge and the trailing edge as text detection, perhaps, extract literal out with threshold process and realize by the concentration or the color of specific character.But, in order to extract the shape of literal accurately, need be used to get rid of literal such as photo edge, printing net-point or the noise various pre-treatments and the reprocessing of part in addition, only extracting the literal shape exactly is not to be easy to.In addition, in the technology of putting down in writing in the patent documentation 1, though extract exactly under the text filed situation no problem, but do not extracting exactly under the text filed situation, exist text filed being programmed in the background layer, non-reversible compression causes the literal breakage, the problem that the visuality of literal worsens.
The present invention finishes in view of such situation, its purpose is to provide by extracting the profile that utilizes maximum black literal in various documents and generates foreground layer, can improve image compressing device, compressed image output device, method for compressing image, computer program and the recording medium of the visuality that is included in the literal in the compressed images.
The image compressing device of first feature of the present invention comprises: detecting color of pixel from original image is the edge of the literal and/or the lines of black, generates the 1st image production part of representing the 1st image at the edge that detects with bianry image; The reversible compression unit that the 1st image that generates is carried out reversible compression; Generate to adjust the 2nd image that is included in the color of pixel in the original image, the 2nd image production part that the concentration difference of color between the pixel that is included in the pixel that comprises in the edge that described the 1st image production part detects and other pixel is reduced in original image; The non-reversible compression unit that the 2nd image that generates is carried out non-reversible compression; The compressed file generating unit of the compressed file that gathers with the 2nd image that generates the 1st image of reversible compression and non-reversible compression.
Of the present invention second is characterized as image compressing device, described the 2nd image production part, the concentration value of concentration that expression is included in the color of pixel at the edge that described the 1st image production part detects be replaced into eight neighborhood territory pixels that are positioned at this pixel around concentration value average of pixel.
The of the present invention the 3rd is characterized as image compressing device, and the concentration value that described the 2nd pixel generating unit will not be included in the pixel in the edge that described the 1st image production part detects carries out smoothly.
The image compressing device of the 4th feature of the present invention also comprises: the low resolution portion that the image before carrying out non-reversible compression is carried out resolutionization.
The image compressing device of the 5th feature of the present invention also comprises: for adjusted color of pixel original image before by described the 2nd image production part, revise the processing of the concentration value of each pixel, make the concentration of each color of pixel become the 1st higher concentration correction portion; With for the 2nd image before the non-reversible compression, revise the processing of the concentration value of each pixel, make the concentration of each color of pixel become the 2nd higher concentration correction portion.
The of the present invention the 6th is characterized as image compressing device, described the 1st image production part calculates the mean value of concentration value of the color depth of the image in the edge that expression is included in detection, and this image compressing device comprises that also the color assignment with the 1st image of reversible compression is the color determination section of the described mean value that calculated.
The image compressing device of the 7th feature of the present invention comprises: aforesaid image compressing device; Reception should be accepted portion by the image that this image compressing device compresses original image; With will output to outside efferent by the compressed file that described image compressing device generates.
The method for compressing image of the 8th feature of the present invention comprises: the step that detects the edge of literal that color of pixel is a black and/or lines from original image; The step of the 1st image at the edge that detects is represented in generation with bianry image; The step that the 1st image that generates is carried out reversible compression; Generate to adjust the 2nd image that is included in the color of pixel in the original image, make in the pixel that original image comprises, comprise the step that the concentration difference of color between the pixel at edge of detection and other pixel is dwindled; The step that the 2nd image that generates is carried out non-reversible compression; With generate will be through the 1st image of reversible compression and the step that is summarised in compressed file together through the 2nd image of non-reversible compression.
In the present invention, the image compressing device of the original image that compression is made of a plurality of pixels detects the black literal in the original image or the edge of lines, the foreground layer (the 1st image) at the detected edge that generation is represented with bianry image, foreground layer is carried out reversible compression, the background layer (the 2nd image) that the concentration difference of the edge of black literal in the original image or lines and other parts has been dwindled in generation carries out non-reversible compression with background layer.The profile of black literal or lines is isolated at edge by detecting black literal or lines and as foreground layer clearly.In addition, by dwindling concentration difference, be reduced in and carry out non-reversible compression and the time result from the noise of compression.
In addition, in the present invention, about being included in the pixel in black literal or the lines, the concentration value by obtaining surrounding pixel average dwindles in the black literal of the periphery of black literal or lines or the concentration difference between lines and the background.
In addition, in the present invention,,, dwindle the variation of the periphery regional concentration in addition of black literal or lines by carrying out the smoothing processing of concentration about the pixel beyond the edge of black literal or lines.
In addition, in the present invention, on the basis of the resolution that has reduced background layer,, the size of compressed file is reduced by carrying out non-reversible compression.
In addition, in the present invention, for the background layer before dwindling the original image before the concentration difference and carrying out non-reversible compression, become the concentration correction of high concentration by making pixel concentration, the image processing when suppressing owing to image compression makes the colour saturation reduction in the image.
In addition, in the present invention, the color of foreground layer that is decided to be the edge of black literal of expression or lines by the mean value with the black concentration in the edge of detected black literal or lines can make the color of black literal in the image of compressed file of expansion or lines become black color near the black in the background layer.
In the present invention, in the image that has stretched compressed file,, and reduce noise, therefore suppressed literal or lines breakage, the visual deterioration, the visuality of raising literal or lines because of compression because the profile of literal or lines is deceived in performance clearly.Owing to improved the visuality of utilizing maximum black literal in various documents, even therefore in the image of compression document, the present invention also can bring into play the visual high of literal, and easily utilizes the outstanding effect of letter again.
Followingly above-mentioned and further purpose and feature are described in detail with reference to accompanying drawing.
Description of drawings
Fig. 1 is the block diagram of built-in function structure of the image processing system of expression execution mode 1.
Fig. 2 is the block diagram of the internal structure of expression compression handling part.
Fig. 3 is the block diagram of other internal structure example of expression compression handling part.
Fig. 4 is the block diagram of other internal structure example of expression compression handling part.
Fig. 5 is the block diagram of other internal structure example of expression compression handling part.
Fig. 6 is the flow chart of the expression foreground layer generating unit processing sequence of carrying out.
Fig. 7 is the concept map of test example at the edge of the black literal of expression.
Fig. 8 is the processing of expression by Filtering Processing and the black pixel of judgement, detects the concept map of the example at the edge of deceiving literal.
Fig. 9 is expression by Filtering Processing and the processing of judging black pixel, has detected the concept map of example at the edge of black literal.
Figure 10 is the determination processing of expression by Filtering Processing and black pixel, has detected the concept map of example at the edge of black literal.
Figure 11 is the ideograph of the example of expression original image.
Figure 12 is the ideograph of the example of expression foreground layer.
Figure 13 is the result of rim detection is only carried out in expression to original image a ideograph.
Figure 14 is the concept map that expression obtains the example of average weighted filter.
Figure 15 is the concept map of the example of expression smoothing filter.
Figure 16 is the concept map that the indicated concentration difference is dwindled the example of the result in the portion.
Figure 17 is the concept map that the indicated concentration difference is dwindled the example of the result in the portion.
Figure 18 is the result's of expression low resolution processing the concept map of example.
Figure 19 is the performance plot of the example of expression look-up table.
Figure 20 is the result of concentration correction is carried out in expression by concentration correction portion the concept map of example.
Figure 21 is the result of concentration correction is carried out in expression by concentration correction portion the concept map of example.
Figure 22 is the ideograph of the example of the result in the expression compression handling part.
Figure 23 is the ideograph of the demonstration example of the expression operation screen that is used to set the method that generates compressed file.
Figure 24 is the block diagram of built-in function structure of the scanning means of the expression image compressing device that comprises execution mode 2.
Figure 25 is the block diagram of internal structure of the image compressing device of the present invention of expression example 3.
Embodiment
Below, specifically describe the present invention according to the drawing of representing embodiment of the present invention.
(execution mode 1)
In execution mode 1, represent that image compressing device of the present invention constitutes the form of the part of the image processing system that forms coloured image.Image processing system 1 is corresponding with compressed image output device of the present invention.Fig. 1 is the block diagram of built-in function structure of the image processing system 1 of expression execution mode 1.Image processing system 1 is the digit apparatus with color copying capability and chromoscan function etc.Image processing system 1 possesses the coloured image input unit 11 (image is accepted portion) that reads coloured image from original copy optically, is connected with to generate and the corresponding view data of coloured image that is read and the color image processing apparatus 2 of compressing image data on coloured image input unit 11.On color image processing apparatus 2, be connected with: according to the color image data of color image processing apparatus 2 generations, the coloured image output device 13 of output coloured image, and the compressing image data that color image forming apparatus 2 is generated sends to the outside, and, receive R-T unit 14 (efferent) from the compressing image data of outside.In addition, on coloured image input unit 11, color image processing apparatus 2, coloured image output device 13 and R-T unit 14, be connected with the operation screen 12 of the operation that receives the user.
Control by not shown CPU (central processing department) in the various processing that image processing system 1 carries out.The CPU of image processing system 1 is by not shown network interface card and LAN (local area network (LAN)) cable, and is connected computer on the communication network and other digit apparatus etc. and carries out data communication.
Secondly, the structure of the each several part of image processing system 1 is described.Operation screen 12 constitutes and comprises: the display part that is used to set the setting indication etc. of the pattern of image processing system 1, the portion that accepts that is used in touch-screen that the operation of indication by the user of the action of control image processing system 1 receive or numerical key etc. and the LCD of the necessary information of expression operation etc.
Coloured image input unit 11 uses the color scanner that possesses CCD optical sensors such as (charge coupled devices) to constitute, to look like to be decomposed into RGB (R: red from the reverberation of the original copy that is recorded in the coloured image on the documentary films such as paper, G: green, B: shades of colour indigo plant), read with optical sensor, be transformed into the analog signal of RGB, to color image processing apparatus 2 outputs.
Color image processing apparatus 2 is for the analog signal of the RGB that imports from coloured image input unit 11, by carrying out image processing described later by A/D transformation component 20, stain correction portion 21, original copy kind judegment part 22, input tonal correction portion 23 and regional separated part 24, the view data that generation is made of the digital signal of RGB (below, be called rgb signal).Below, the intensity of rgb signal is expressed as R, G, B respectively.
In addition, color image processing apparatus 2 is for the rgb signal of regional separating treatment portion 24 outputs, remove portion 26, space filtering handling part 27, output tonal correction portion 28, tone rendering handling part 29 execution image processing described later by generate background color by colour correction portion 25, black, generation is by CMYK (C: viridescent, M: magenta, Y: Huang, K: black) view data that digital signal constitutes is as 13 outputs of data flow coloured image output device.Wherein, color image processing apparatus 2 also can be before coloured image output device 13 output coloured images, hard disk etc. as the storage part 30 of Nonvolatile memory devices in the structure of temporary transient storing image data.
Coloured image output device 13 is based on the view data by color image processing apparatus 2 input, by duplicate according to heat, electronics is taken or mode such as ink-jet forms coloured image, the output coloured image in documentary films such as paper.Wherein, replace coloured image output device 13, image processing system 1 also can possess the monochrome image output device that forms monochrome image output in documentary film.In this case, color image processing apparatus 2 is transformed in the view data with coloured image on the basis of view data of monochrome image, to monochrome image output device output image data.
In addition, color image processing apparatus 2 is by carrying out method for compressing image of the present invention described later to the rgb signal of regional separating treatment portion 24 outputs by compression handling part 3, generation comprises the compressed file with the data of Image Data Compression, to R-T unit 14 output compressed files.Compression handling part 3 plays the effect of image compressing device of the present invention.Wherein, color image processing apparatus 2 also can be the structure that temporarily was stored in before the output of R-T unit 14 compressed files in the storage part 30.
R-T unit 14 is made of network interface card or modulator-demodulator etc., can be connected with not shown communication networks such as common return net, LAN or internet, according to communication means, via communication network, to outside transmitting-receiving compressed file by fax or Email etc.For example, in operation screen 12, under the situation of selecting scan to e-mail pattern, R-T unit 14 adds in the Email compressed file to the transmit leg transmission of setting to.
Wherein, during the transmission of being faxed by R-T unit 14, the CPU of image processing system 1 communicates formality with R-T unit 14 with the other side, when guaranteeing for the state that can send, after compressed file being implemented the necessary processing such as change of compressed format, in turn send to the other side.
In addition, when R-T unit 14 is received faxes, when the CPU of image processing system 1 communicates formality with R-T unit 14, receive the compressed file that sends from the other side, import the compressed file that receives to color image processing apparatus 2 by R-T unit 14.2 pairs of compressed files that receive of color image processing apparatus stretch processing by not shown stretching, extension handling part.In addition, 2 pairs of view data that obtain through the stretching, extension compressed file of color image processing apparatus, be rotated processing and/or resolution conversion processing etc. by not shown handling part as required, output tonal correction portion 28 exports tonal correction, and tone rendering handling part 29 carries out tone rendering to be handled.And then color image processing apparatus 2 is exported the view data of having implemented various image processing to coloured image output device 13, and coloured image output device 13 is according to view data output coloured image.
Then, image processing in the color image processing apparatus 2 is described.A/D transformation component 20 receives from the analog signal of coloured image input unit 11 to the RGB of color image processing apparatus 2 inputs, the analog signal of RGB is transformed to the rgb signal of digital signal, to stain correction portion 21 output rgb signals.
21 pairs of stain correction portions are removed the processing of the various distortion that produce from the rgb signals of A/D transformation component 20 inputs in illuminator, imaging system and the camera system of coloured image input unit 11.Stain correction portion 21 is then removed the rgb signal of distortion to 22 outputs of original copy kind judegment part.
Original copy kind judegment part 22 will be transformed into the concentration signal of the various colour saturations of expression RGB from the reflectivity signals of the RGB of stain correction portion 21 input, discern the original copy kind discriminating processing of the original copy pattern of literal, printing photo or photographic pape photo etc.Pass through to operate screen 12 the user and manually set under the situation of original copy kind discriminating processing, the original copy kind judegment part 22 input tonal correction portion 23 of one-level is backward exported the rgb signal that has been transformed into the concentration signal of representing the various colour saturations of RGB.The result of original copy kind discriminating processing obtains reflection in the image processing of back one-level.
Input tonal correction portion 23 pairs of rgb signals carry out color balance adjustment, base concentration remove and the image quality adjustment of the adjustment of contrast etc. is handled.Then, input tonal correction portion 23 has carried out the rgb signal of handling to 24 outputs of regional separating treatment portion.
Zone separating treatment portion 24 will be from each pixel separation the image that the rgb signal of input tonal correction portion 23 inputs is represented for deceiving the some of character area, dot area or photo zone.As the processing that is used to detect black character area, regional separating treatment portion 24 carries out processing at edge that detects literal in aftermentioned ground and the processing of judging black pixel.In addition, regional separating treatment portion 24 generates background color to compression handling part 3, black and removes the regional identification signal which zone portion 26, space filtering handling part 27 and tone rendering handling part 29 each pixel of output expression belong to based on separating resulting.Zone separating treatment portion 24 and then to colour correction portion 25 and 3 outputs of compression handling part rgb signal from 23 inputs of input tonal correction portion.Wherein, regional separating treatment portion 24 also can be for to the form of compression handling part 3 output area identification signals.
Black generates background color and removes portion 26 based on the CMY signal of importing from colour correction portion 25, carries out generating the black generation processing of black (K) signal and deducting the new CMY Signal Processing of K signal generation that obtains from original C MY signal black generates from the CMY signal.Its result, the CMY signal transformation of three looks is the digital signal of CMYK four looks (below, be called the CMYK signal).Black generates background color and removes the CMYK signal that portion 26 is then come by CMY signal transformation to 27 outputs of space filtering handling part.
Generate background color as black and remove the black that portion 26 carries out and generate an example of handling, the general use based on profile black carried out the method that black generates.In the method, if establish the input-output characteristic of contour curve is y=f (x), the data of input are C, M, Y, the data of output are C ', M ', Y ', K ', UCR (background color is removed) rate is α (0<α<1), then generates background color by black and removes portion 26 and handle C ', the M ', Y ', the K ' that obtain and represent with following formula (1)~formula (4).
K’=f(min(C,M,Y))……(1)
C’=C-αK’……(2)
M’=M-αK’……(3)
Y’=Y-αK’……(4)
At this, UCR leads α (0<α<1) and represents that the overlapping part of CMY is replaced into K which kind of degree CMY has been cut down.Formula (1) is illustrated in each signal strength signal intensity of CMY, the K signal that generates corresponding to the signal strength signal intensity of minimum.
27 pairs of images of removing the CMYK signal indication of portion's 26 inputs from black generation background color of space filtering handling part, based on regional identification signal from 24 inputs of regional separating treatment portion, handle the correction spatial frequency characteristic by the space filtering of being implemented by digital filter, the fuzzy or graininess that improves image worsens.For example, for the zone that is separated into literal in regional separating treatment portion 24, space filtering handling part 27 is in order to improve the reproducibility of literal, uses the big filter of the amount of emphasizing of radio-frequency component to carry out space filtering and handles.In addition, for the zone that is separated into the site in regional separating treatment portion 24, space filtering handling part 27 is used to remove the low-pass filtering treatment of input site composition.Space filtering handling part 27 is the CMYK signal after 28 outputs of output tonal correction portion are handled then.
Output tonal correction portion 28 is for the CMYK signal from 27 inputs of space filtering handling part, make corresponding with the characteristic of coloured image output device 13, carry out handling based on the output tonal correction of dot area percentage, the CMYK signal after the output tonal correction is handled is to 29 outputs of tone rendering handling part.
Tone rendering handling part 29 is for the CMYK signal from 28 inputs of output tonal correction portion, based on the regional identification signal from 24 inputs of regional separating treatment portion, carries out and regional corresponding semi-tone is handled.For example, for the zone that is separated into literal in regional separating treatment portion 24, tone rendering handling part 29 carries out binaryzation or many-valuedization processing by being suitable for reproducing the high-resolution reseau of radio-frequency component.In addition, for the zone that is separated into the site in regional separating treatment portion 24, tone rendering handling part 29 carries out the processing of binaryzation or many-valuedization by the reseau of paying attention to tone rendering.The view data of tone rendering handling part 29 after 13 outputs of coloured image output device are handled.
Then, the structure of compression handling part 3 and the processing that compression handling part 3 carries out are described.Compression handling part 3 is based on the view data that is made of rgb signal from 24 inputs of regional separating treatment portion, with method for compressing image generation compressed file of the present invention.
The original image that is input to the pictorial data representation of the coloured image that compresses handling part 3 constitutes by being configured to rectangular a plurality of pixels.Compression handling part 3 generates foreground layer (the 1st image) and background layer (the 2nd image) from original image, and foreground layer is transformed into bianry image, for example uses the reversible compression foreground layer of MMR, for example with JPEG background layer is carried out non-reversible compression.Compression handling part 3 then will be through the foreground layer of reversible compression and through the background layer of non-reversible compression with to be used for the stretching, extension information that they stretch the view data that forms coloured image is gathered be 1 file.This document is a compressed file.In addition, as this stretching, extension information, the information of use expression compressed format and the colored form of index (below, be called the IC table) etc.Such compressed file with directly compare to the situation of the Image Data Compression of coloured image or to the situation of foreground layer and background layer compression, document size is little, but also suppresses the deterioration of image quality.In addition and since can possess separately respectively a kind of for bianry image reversible compression method and for the non-reversible compression method of background layer, therefore do not need to prepare the compression method more than three kinds.
Fig. 2 is the block diagram of the internal structure example of expression compression handling part 3.Compression handling part 3 possess foreground layer generating unit 31 (the 1st image production part), the compression foreground layer of the foreground layer that generates the expression literal reversible compression unit 32, carry out non-reversible compression unit 35 and compressed file generating unit 33 (color determination section) that the concentration difference of processing of the concentration difference of foreground layer in the downscaled images and other parts is dwindled portion 34 (the 2nd image production part), compressed background layer.Be input to foreground layer generating unit 31 and concentration difference is dwindled portion 34 from the regional identification signal of regional separated part 24 and by the view data that rgb signal constitutes.Foreground layer generating unit 31 is extracted the signal of black character area out from regional identification signal, by generating corresponding to the pixel at the edge of the black literal bianry image with two value representations, generation foreground layer.Foreground layer generating unit 31 is exported the foreground layer that generates to reversible compression unit 32,32 pairs of foreground layers from 31 inputs of foreground layer generating unit of reversible compression unit carry out reversible compression.
Concentration difference is dwindled portion 34 based on the signal that is included in the black character area in the regional identification signal, in the pixel that in original image, comprises, specified packet is contained in the pixel in the black character area, dwindles the processing of the concentration difference of pixel in the black character area and other pixel.Specifically, concentration difference is dwindled portion 34 for the pixel on the black literal edge, makes the Filtering Processing that concentration difference reduces between near the pixel the edge.In addition, concentration difference is dwindled portion 34 for the pixel that is not included in the black character area, carries out the smoothing processing of concentration.Concentration difference dwindle portion 34 for original image by carrying out such Filtering Processing and smoothing processing, the generation background layer, with background layer to non-reversible compression unit 35 outputs.Non-reversible compression unit 35 carries out non-reversible compression with the compression ratio higher than reversible compression unit 32 to background layer.
Wherein, at the regional identification signal of the signal that comprises the black character area of expression is not under the situation of the form of regional separating treatment portion 24 inputs, foreground layer generating unit 31 detects the processing of black character area by processing that detects the literal edge and the processing of judging black pixel.Make the situation of color image processing apparatus 2 actions under this form and the setting of not using regional separating treatment portion 24 or do not possess the color image processing apparatus 2 of regional separating treatment portion 24 suitable.Foreground layer generating unit 31 is carried out general rim detection Filtering Processing such as Ru Beier (Sobell) rim detection filtering as the processing at the edge that detects literal.In addition, as the processing that detects black pixel, foreground layer generating unit 31 is in turn selected the pixel in the original image, under the situation of R=G=B<TH1, judges and gazes at the processing of pixel for black pixel.At this, TH1 is used to judge gaze at the threshold value that color of pixel is the rgb signal intensity of black, is the value that is stored in advance in the foreground layer generating unit 31.In addition, foreground layer generating unit 31 is as the processing that detects black pixel, | R-G|, | G-B| and | B-R| is respectively below threshold value TH2, and each of R, G and B also can be judged and gaze at the processing that pixel is black pixel all under the situation below the TH1.At this, the value of TH2 is littler than TH1, and is stored in advance in the foreground layer generating unit 31.The foreground layer generating unit 31 that detects black character area is dwindled the signal of the black character area of portion's 34 output expressions to concentration difference, concentration difference is dwindled portion 34 based on the signal of deceiving character area from the expression of foreground layer generating unit 31 inputs, carries out the processing that specified packet is contained in the pixel in the black character area.
Fig. 3~Fig. 5 is the block diagram of other internal structure example of expression compression handling part 3.Fig. 3 represents to make the form of the compression handling part 3 of background layer low resolutionization.Compression handling part 3 also possesses the low resolution portion 36 that makes the background layer low resolutionization except that foreground layer generating unit 31, reversible compression unit 32, compressed file generating unit 33, concentration difference are dwindled portion 34 and non-reversible compression unit 35.Concentration difference is dwindled portion 34 to low resolution portion 36 output background layers, and low resolution portion 36 will be transformed to low-resolution image from the background layer that concentration difference is dwindled portion's 34 inputs, to the background layer of non-reversible compression unit 35 output low resolutionization.35 pairs of background layers from 36 inputs of low resolution portion of non-reversible compression compression unit carry out non-reversible compression, to the non-reversible packed data of compressed file generating unit 33 outputs.Foreground layer generating unit 31, reversible compression unit 32 and compressed file generating unit 33 are carried out processing same as described above.By carrying out the low resolutionization of background layer, can further reduce the data size of compressed file.
Fig. 4 represents to carry out the form of compression handling part 3 of the concentration correction of background layer.Compression handling part 3 also possesses the concentration correction portion 37 (the 1st concentration correction portion) and 38 (the 2nd concentration correction portions) of the concentration correction of carrying out background layer except that foreground layer generating unit 31, reversible compression unit 32, compressed file generating unit 33, concentration difference are dwindled portion 34 and non-reversible compression unit 35.View data and regional identification signal from regional separating treatment portion 24 are input to foreground layer generating unit 31 and concentration correction portion 37.Concentration correction portion 37 handles the concentration correction of the concentration higher concentrationization of the pixel that comprises in the original image.Concentration correction portion 37 dwindles the view data that concentration correction has been carried out in portion's 34 outputs to concentration difference, and concentration difference is dwindled portion 34 by the view data generation background layer from 37 inputs of concentration correction portion, to 38 outputs of concentration correction portion.Concentration correction portion 38 makes the concentration correction processing of dwindling the concentration higher concentrationization of the pixel that comprises the background layer of portion's 34 inputs from concentration difference, has carried out the background layer of concentration correction to non-reversible compression unit 35 outputs.35 pairs of background layers from 38 inputs of concentration correction portion of non-reversible compression unit carry out non-reversible compression, to the non-reversible packed data of compressed file generating unit 33 outputs.Foreground layer generating unit 31, reversible compression unit 32 and compressed file generating unit 33 are carried out processing same as described above.
Fig. 5 represents to carry out the form of the compression handling part 3 of the low resolutionization of background layer and concentration correction.Compression handling part 3 also possesses low resolution portion 36, concentration correction portion 37 and 38 except that foreground layer generating unit 31, reversible compression unit 32, compressed file generating unit 33, concentration difference are dwindled portion 34 and non-reversible compression unit 35.View data and regional identification signal from regional separated part 24 are input to foreground layer generating unit 31 and concentration correction portion 37, and concentration correction portion 38 carries out the processing of concentration correction, and the view data that will carry out concentration correction outputs to concentration difference and dwindles portion 34.Concentration difference is dwindled portion 34 by the view data generation background layer from 37 inputs of concentration correction portion, to low resolution portion 36 output background layers.Low resolution portion 36 will be transformed into low-resolution image from the background layer that concentration difference is dwindled portion 34 input, to the background layer of concentration correction portion 38 output low resolutionization.38 pairs of background layers from 36 inputs of low resolution portion of concentration correction portion carry out the processing of concentration correction, have carried out the background layer of concentration correction to non-reversible compression unit 35 outputs.35 pairs of background layers from 38 inputs of concentration correction portion of non-reversible compression unit carry out non-reversible compression, to the non-reversible packed data of compressed file generating unit 33 outputs.Foreground layer generating unit 31, reversible compression unit 32 and compressed file generating unit 33 are carried out processing same as described above.
Then, the processing of carrying out in the foreground layer generating unit 31 is described.Fig. 6 is the flow chart of the order of the processing carried out of expression foreground layer generating unit 31.Foreground layer generating unit 31 based in the regional identification signal of regional separating treatment portion 24 inputs, remarked pixel is included in the character area signal the character area, detects the edge (S1) of the black literal that comprises from the original image of the pictorial data representation of regional separating treatment portion 24 inputs.Fig. 7 is the concept map that detects example at the edge of the black literal of expression.The form example of Fig. 7 (A) expression regional identification signal.At example of Fig. 7 (A) expression, regional identification signal is the signal of 3 bits, and the bit of left end is represented the zone of black literal, and the bit of central authorities is represented dot area, and the bit of right-hand member is represented the photo zone.Each bit is if 1, and then pixel packets is contained in the zone that its bit represents, bit is if 0, and then pixel is not included in the zone that its bit represents.Fig. 7 (B) expression makes the example of regional identification signal corresponding to each pixel in the original image.The bit that is included in expression character area in the pixel of original image by extraction is 1 pixel, detects the black character area in the original image.And then, in the pixel of extracting out, be the pixel of character area by removing two neighbours, neighbours territory or eight neighborhoods, remaining pixel becomes the pixel corresponding with the edge of deceiving literal, detects the edge of black literal.Fig. 7 (C) expression is with the pixel corresponding with the edge of black literal in each pixel shown in 1 presentation graphs 7 (B), with 0 state of representing other pixel.By step S1, can obtain the two-value mask at the black literal edge of the expression shown in Fig. 7 (C).
Wherein, do not have from regional separating treatment portion 24 under the situation of form of input area identification signal, foreground layer generating unit 31 is in step S1, and the processing at the edge by detecting literal and the processing of judging black pixel detect the processing at the edge of black literal.Fig. 8~Figure 10 is by Filtering Processing and judges the processing of deceiving pixel, detects the concept map of the example at the edge of deceiving literal.Fig. 8 (A) represents the image with the concentration of 0~255 each pixel of value representation.Fig. 8 (B) expression two types 3 * 3 Ru Beier edge in length and breadth detects filter.Fig. 9 (A) expression is to the image of Fig. 8 (A) expression, and two kinds of Ru Beier edges that will be suitable for Fig. 8 (B) expression detect the result who takes absolute value after the filtering results added of filter.Become 0 with the pairing value of pixel of part beyond the literal edge, the pixel corresponding with bigger value is detected as the literal edge.Foreground layer generating unit 31 is 1 by making the value bigger than predetermined threshold value then, and the following value of threshold value is 0, carries out the binary conversion treatment that detects the result at edge.Fig. 9 (B) expression is 500 with threshold value, carries out the result of binary conversion treatment.The result of binaryzation, the pixel corresponding to 1 is detected as the literal edge.Foreground layer generating unit 31 then carry out for the result's 1 of binaryzation the corresponding pixel of value, take a decision as to whether black pixel, under the situation that is being black pixel, make with the corresponding value of pixel be 1, under the situation that be not black pixel, making with the corresponding value of pixel is 0 processing.As the processing of judging black pixel, as mentioned above, whether be R=G=B<TH1, perhaps, | R-G|, | G-B| and | whether B-R| respectively below TH2, and each all processing below TH1 of R, G and B.When Figure 10 represents as TH1=60 and TH2=10, judge the result of black pixel.Be the final pixel that is detected as the edge of black pixel with the corresponding pixel of 1 value among Figure 10.
Foreground layer generating unit 31 is after step S1 finishes, calculate the summation of the concentration value in the pixel that detects as the edge of deceiving literal, the pixel count at the edge of the black literal of counting by remove the summation of concentration value with pixel count, calculates the mean value (S2) of the black concentration at the edge of deceiving literal.As the concentration value of pixel, for example use the mean value of the value of R, G and B.In compression handling part 3, the mean value of the black concentration that the calculates black information as the black concentration of the black literal of expression is used.Thus, can be to replace black literal edge near the concentration of the black concentration of compressing preceding original image as far as possible.In addition,, also can be redefined for fixed value, and omit the form of step S2 deceiving concentration in order to simplify processing.
Foreground layer generating unit 31 will be then by with being 1 with the concentration value of the corresponding pixel in edge of detected black literal, and the concentration value of other pixel is 0 method, generates bianry image from original image, generates foreground layer (S3).The foreground layer generating unit 31 that generates foreground layer is to reversible compression unit 32 output foreground layer and black informations, end process.
Figure 11 is the ideograph of the example of expression original image.Be included in " TEST " literal of 5 sections in the original image that Figure 11 represents, the literal of the 1st section, the 3rd section and the 5th section is black literal, and the literal of the 2nd section and the 4th section is formed by colour.In addition, the background of the 1st section, the 2nd section and the 3rd section literal is a white, and the background of the 4th section and the 5th section literal is colored.Figure 12 is the ideograph of the example of expression background layer, and the original image that expression is represented from Figure 11 detects the example of the edge generation bianry image of black literal.Black part is that the concentration value of pixel is 1 part among the figure, and white part is that the concentration value of pixel is 0 part among the figure.Wherein, concentration value 0 expression is transparent.As shown in figure 12, be included in the literal of Figure 11, only detect the edge of black literal, generating bianry image is foreground layer.
Figure 13 is an ideograph of representing original image has only been carried out the result of rim detection, and the original image that expression is represented from Figure 11 detects the result at the edge of literal.Under the situation of the form that does not have the input area identification signal from regional separating treatment portion 24, foreground layer generating unit 31 obtains result shown in Figure 13 by detecting the edge of literal.In this stage, as shown in figure 13, detect the edge of black literal and with the edge of the literal of colored formation.Foreground layer generating unit 31 by judging the processing of black pixel, obtains bianry image as shown in Figure 12 for the result of as shown in Figure 13 rim detection.
Then, illustrate that concentration difference dwindles the processing in the portion 34.Concentration difference is dwindled portion 34 and is selected to be included in pixel in the original image successively, judges the pixel of gazing in the edge whether pixel be included in black literal.Specifically, concentration difference is dwindled portion 34 by the pixel in the edge that carries out the processing identical with above-mentioned foreground layer generating unit 31, judge gazing at pixel and whether be included in black literal.Under the situation of the pixel in gazing at the edge that pixel is included in black literal, concentration difference is dwindled the average weighted Filtering Processing that portion 34 obtains the concentration value of the pixel around more being positioned at of eight neighborhood territory pixels of gazing at pixel.Figure 14 is the concept map that expression obtains the example of average weighted filter.By using this filter, the weighted average of the concentration value of the pixel around eight neighborhood territory pixels that can obtain gazing at pixel more are positioned at can reduce the concentration difference between pixel and its pixel on every side in the black character area.
Gazing at pixel is not to be included under the situation of the pixel in the black character area, and concentration difference is dwindled portion 34 and used smoothing filter to carry out smoothing processing.At this, the smoothing processing of carrying out is made as the smoothing processing that is attenuated to the moir degree that the site does not take place.Figure 15 is the concept map of the example of expression smoothing filter.By using this smoothing filter, gazing at pixel and gazing between the pixel of eight neighborhoods of pixel and carry out smoothing processing.Wherein, the filter factor of representing among Figure 14 and Figure 15 is an example, is not limited to this example.
Figure 16 and Figure 17 are the concept maps that the indicated concentration difference is dwindled the example of the result in the portion 34.The example of Figure 16 (A) expression original image, Figure 16 (B) expression detect the result at the edge of black literal from the original image of Figure 16 (A) expression.Figure 17 represents the original image for Figure 16 (A) expression, uses the Filtering Processing of the filter that Figure 14 and Figure 15 represent and the result of smoothing processing.Wherein, in this example, all calculate as R=G=B=255 in the zone in the not shown outside of image.By carrying out such Filtering Processing, the concentration difference on every side at black literal edge is reduced, can suppress because the ringing noise (RingingNoise) that non-reversible compression takes place.In addition,, near the zone beyond the edge, can suppress to take place the moir of ringing noise and site, also help further improving compression ratio by smoothing processing.The result that Figure 17 represents is equivalent to background layer.Concentration difference is dwindled portion 34 by obtaining result as shown in Figure 17, the generation background layer, and with the background layer output that is generated.
Then, processing in the low resolution portion 36 is described.Low resolution portion 36 for concentration difference dwindle background layer that portion 34 generates by sample merely, neighborhood analysis (NearestNeighbor), bilinearity method (Bi-linear) or three interpolation method interpolations such as (Bicubic) handle, and carry out the low resolution processing of background layer.Figure 18 is the concept map of the example of expression low resolution result.The image of representing among Figure 18 is that the background layer that will represent among Figure 17 makes resolution be reduced to the result of 1/2 resolution with the bilinearity method.At this, as the initial point of bilinearity method since with the center of 2 * 2 pixels as initial point, therefore the mean value (rounding up below the decimal point) of 2 * 2 pixels is the result of bilinearity method.As shown in figure 18, the pixel count by the low resolution background layer reduces the reduction data volume.Be included in the trickle shape of literal in the original image owing to can in foreground layer, reproduce, even it is therefore reduce the resolution of background layer, also less to the influence of the reproducibility of image.Thereby, can not cause the reproducibility of image to worsen, can reduce the compressed file size.
Then, processing in concentration correction portion 37 and 38 is described.Owing to dwindle Filtering Processing in the portion 34, low resolution processing in the low resolution portion 36 and the non-reversible compression in the non-reversible compression unit 35 and handle by carrying out concentration difference, the concentration of the general image of background layer is corrected for more even, so the contrast reduction of background layer.Thereby, dwindle before the Filtering Processing of portion 34 in concentration difference and before the non-reversible compression of non-reversible compression unit 35 handles, make the pixel concentration of background layer become the concentration correction of high concentration. Concentration correction portion 37 and 38 is stored in the look-up table of one dimension respectively, handles with the concentration value of each pixel in the look-up table changing background layer.
Figure 19 is the performance plot of the example of expression look-up table.Look-up table in Figure 19 (A) indicated concentration correction portion 37, the look-up table in Figure 19 (B) indicated concentration correction portion 38.Transverse axis among the figure represents that the concentration value imported, the longitudinal axis represent the concentration value of exporting.The concentration value of each pixel of background layer as shown in figure 19, with the concentration value of the corresponding conversion of the concentration value output of input.As shown in figure 19, look-up table is set on the value ground that the concentration value of input is transformed into below present concentration value.Because the value of the concentration value color of small pixel more is dense more, so the concentration of the pixel of background layer is by high concentrationization.In the example of Figure 19 (A) expression, the concentration value 0~30 of input is transformed into concentration value 0, the concentration value 30~64 of input is transformed into deducts 30 concentration value from input.In addition, the concentration value of input is the relation that changes input and output between 65~100 time linearly, makes concentration value=100 of input, the concentration value of output=100.In addition, the concentration value 100~255 to input carries out conversion with input=output.In the example of Figure 19 (B) expression, the concentration value 0~20 of input is transformed into concentration value 0, the concentration value of input is the relation that changes input and output between 20~128 time linearly, makes concentration value=128 of input, the concentration value of output=128.In addition, the concentration value 128~255 of input is transformed into input=output.For the background layer that in concentration correction portion 37, has carried out concentration correction, owing to before non-reversible compression, carry out the concentration correction in the concentration correction portion 38 once more, therefore as shown in figure 19, the concentration correction of the degree of high concentrationization in can specific concentration correction portion 37 is slightly good.Wherein, the look-up table that Figure 19 represents is an example, is not limited to this example.In addition, the look-up table that uses in concentration correction portion 37 and 38 both can look like different mutually as shown in Figure 19, also can be identical.Under the identical situation of look-up table, can reduce needed memory capacity in the compression handling part 3.Under the different situation of look-up table, can carry out reducing corresponding finer concentration adjustment with each concentration of handling of compression handling part 3.
The concept map of the example of the result of Figure 20 and Figure 21 concentration correction that to be expression implemented by concentration correction portion 37 and 38 is illustrated in the result of the compression handling part 3 that Fig. 5 represents.Figure 20 (A) expression is carried out the result that concentration correction is handled to the original image of Figure 16 (A) expression in concentration correction portion 37.Compare with Figure 16 (A), the concentration value of each pixel is transformed into the denseer value of color (littler value).Figure 20 (B) expression is dwindled the background layer that portion 34 generates according to the result of Figure 20 (A) expression in concentration difference.Compare with the background layer that Figure 17 represents, background layer is by higher concentrationization.Figure 21 (A) expression is to the background layer of Figure 20 (B) expression, the result who carries out the low resolution processing in low resolution portion 36.In addition, Figure 21 (B) expression is carried out the result that concentration correction is handled to the background layer of Figure 21 (A) expression in concentration correction portion 38.Compare with the background layer that Figure 18 represents, the carrying out of Figure 21 (B) expression the concentration value of each pixel of background layer of concentration correction further reduce, become high concentration.Therefore, even carried out non-reversible compression by non-reversible compression unit 35 afterwards, also can make contrast keep higher.
The processing of the each several part by aforesaid compression handling part 3 generates foreground layer and background layer from original image, and foreground layer, black information and background layer are input to compressed file generating unit 33.Compressed file generating unit 33 generates compressed file by being summarised in the file by the foreground layer of reversible compression and by the background layer of non-reversible compression and stretching, extension information.This stretching, extension information comprises that the black concentration of the black literal that black information is represented is decided to be the information of the black color of pixel that is included in foreground layer.Can make the color of the black literal in the image of compressed file of stretching, extension become black color, can suppress the gap of black concentration between the profile of background layer and black literal near the black in the background layer.
Figure 22 is the ideograph of the example of the result in the expression compression handling part 3, the result of the image processing that the original image that expression is represented Figure 11 carries out.Figure 22 (A) represents background layer, and Figure 22 (B) represents foreground layer.Background layer is and original image image much at one that foreground layer is the bianry image at the edge of the black literal of expression.The image of compressed file is stretched in Figure 22 (C) expression, becomes foreground layer and background layer are carried out the laminated image that becomes.In Figure 22,, in fact, use black performance near the concentration of the black text color of background layer though emphasized to describe the edge of black literal.Therefore the profile of black literal can show clearly owing to use by the performance of the foreground layer of reversible compression, deceives the visual high of literal.In addition, because the concentration difference of black literal and other parts further diminishes, the noise that takes place when therefore being suppressed at the non-reversible compression of background layer suppresses to deceive the deterioration of the visuality of literal.About literal with colored formation, the processing by dwindling concentration difference and the processing of non-reversible compression though bleeding takes place, cause hazy sensations, but, can improve visuality by carrying out the processing of the concentration correction of concentration correction portion 37 and 38 under the more suitable condition.
Image processing system 1 also can be can be in the processing of the compressed file that generates the visuality that improves black literal with the present invention and the form of selecting with the processing of method generation compressed file commonly used.Under the form that can select, image processing system 1 becomes the operation that receives from the user by operation screen 12, sets the structure of the method that generates compressed file.Figure 23 is the ideograph of the demonstration example of the expression operation screen 12 that is used to set the method that generates compressed file.Operation shields 12 as shown in figure 23, and display setting picture on display part receives the operation from the user in acceptance division.Become when generating compressed file compression ratio from low compression, the structure selected in compression and high compression three kinds, the example of compressing during expression is selected among Figure 23 (A).The example of high compression is selected in expression in Figure 23 (B), under the situation of selecting high compression, and then can specify the review button of " paying attention to black literal ".In setting picture by so that specify " paying attention to black literal ", set the processing that generates the compressed file of the visuality that improves black literal with the present invention.The document size that reduces compressed file in hope improves under the situation of the visuality of deceiving literal, can generate compressed file with the present invention.Under situation about need not compress with high compression rate, perhaps, be not only black literal, also inferior the situation that the visuality of colored text also improves to a certain degree, can use usual way to generate compressed file.
In addition, image processing system 1 also can be the form that the image based on the compressed file that generates in compression handling part 3 is formed at coloured image output device 13.The image processing system 1 of this form is for the compressed file of compressed file that generates in compression handling part 3 or storage in storage part 30, by stretching processing by not shown stretching, extension handling part, stretch foreground layer and background layer, foreground layer and background layer are carried out laminated one-tenth.Color image processing apparatus 2 carries out necessary image processing for foreground layer and background layer being carried out the view data that laminated one-tenth obtains, to 13 outputs of coloured image output device.Coloured image output device 13 generates coloured image based on view data.
As explained in detail above, in the present invention, when compression comprises the original image of black literal, detect the edge of the black literal in the original image, the foreground layer at detected edge is represented in generation with bianry image, foreground layer is carried out reversible compression, generate the background layer of the concentration difference of the edge reduce the black literal in the original image and other parts, background layer is carried out non-reversible compression.By not being to detect the edge of the whole but black literal of literal as foreground layer, will deceive literal and be programmed in the background layer, breakage can not take place by non-reversible compression literal, thereby be easy to from the background separation literal.The profile of black literal owing to use by the foreground layer of reversible compression is represented, therefore can conclusively show, and has improved the visuality of the black literal in the image of the compressed file that stretches.Owing to improved the visuality of utilizing maximum black literal in various documents, therefore in the image of compression document, the visuality of literal also uprises, and can easily utilize document again.
In addition, in the present invention, for the concentration difference in the downscaled images, for the pixel in the edge that is included in black literal, the average weighted Filtering Processing of the concentration value by obtaining surrounding pixel, at the periphery of black literal, the edge of black literal fogs, and dwindles the concentration difference between black literal and the background.In addition, about the pixel beyond the edge of black literal,, dwindled the change of concentration in the zone beyond the periphery of black literal by carrying out the smoothing processing of concentration.Suppress the concentration difference of black literal and background by periphery, the ringing noise of generation in the time of can being suppressed at the non-reversible compression of background layer at black literal.In addition, by the change of concentration in the zone beyond the periphery that dwindles black literal, in the time of can suppressing the non-reversible compression of background layer, the zone beyond the periphery of black literal, generation ringing noise and Moire fringe.Like this, owing to can suppress to result from the noise of image compression, therefore can suppress to stretch the deterioration of the black literal in the image of compressed file.
(execution mode 2)
In execution mode 1, represented that image compressing device of the present invention becomes the form of the part of image processing system, and in execution mode 2, represented that image compressing device of the present invention becomes the form of the part of scanning means.Scanning means is corresponding with image compression output device of the present invention.Figure 24 is the block diagram of built-in function structure of the scanning means of the expression image compressing device that comprises execution mode 2.Scanning means possesses the coloured image input unit 11 that reads coloured image optically, is connected with image processing apparatus 4 on coloured image input unit 11.On image processing apparatus 4, connect personal computer not shown main devices such as (PC) via telecommunication cable or communication network.Attended operation screen 12 on coloured image input unit 11 and image processing apparatus 4.
Coloured image input unit 11 carries out the processing identical with execution mode 1, and the analog signal of the RGB of the coloured image that reads is outputed to coloured picture as processing unit 4.Image processing apparatus 4 becomes and will be transformed into digital signal A/D transformation component 20 from the analog signal of coloured image input unit 11 input, forms the structure that the order according to stain correction portion 21, original copy kind judegment part 22, input tonal correction portion 23, regional separating treatment portion 24, compression handling part 3 transmits.Compression handling part 3 is corresponding with image compressing device of the present invention.The structure of A/D transformation component 20, stain correction portion 21, original copy kind judegment part 22 and output tonal correction portion 23 is identical with example 1.Zone separating treatment portion 24 will output to compression handling part 3 from the rgb signal of input tonal correction portion 23 inputs.
The structure of compression handling part 3 is identical with execution mode 1, compression handling part 3 is from the view data of regional separating treatment portion 24 inputs as rgb signal, by carrying out the of the present invention method for compressing image identical, generate the compressed file of the view data that compression imports with execution mode 1.In addition, compression handling part 3 is exported the compressed file that is generated to not shown main device.Main device receives the compressed file of image processing apparatus 4 outputs, and the storage, compressed file of carrying out compressed file is to the transmission of outside or based on the processing of the image output of compressed file etc.Wherein, also can be to replace coloured image input unit 11 to use the form of digital camera.
As mentioned above, in the present embodiment, also identical with execution mode 1, when compression comprises the original image of black literal, the edge that detects black literal is as foreground layer, the edge by dwindling the black literal in the original image and the concentration difference of other parts be layer as a setting, and the visuality of the black literal in the image that stretches compressed file is improved.Therefore, the scanning means of present embodiment makes the compressed file of the visuality raising of literal in the time of can being created on reproduced image in the size that reduces compressed file.
(execution mode 3)
In execution mode 3, expression uses the general calculation machine to realize the form of image compressing device of the present invention.Figure 25 is the block diagram of internal structure of the image compressing device of the present invention 5 of expression execution mode 3.The image compressing device 5 of present embodiment is made of all-purpose computers such as PC or server units, possesses RAM52 that CPU51, the storage of carrying out computing be accompanied by the temporary transient information that computing takes place, reads storage parts 54 such as the drive portion such as CD-ROM drive 53 of information and hard disk from recording mediums 6 of the present invention such as CDs.CPU51 makes drive portion 53 read computer program 61 of the present invention from recording medium 6 of the present invention, and the computer program 61 that is read is stored in the storage part 54.Computer program 61 downloads to RAM52 from storage part 54 as required, and based on the computer program of downloading 61, CPU51 carries out necessary processing to image compressing device 5.
In addition, image compressing device 5 possesses input according to the input part 55 of the keyboard of the information such as various processing indications of user operation or indicating equipment etc. with show the display parts such as LCD 56 of various information.And then image compressing device 5 possesses sending part 57 that can be connected with not shown external communication network and the acceptance division 58 that is connected with the external input device 62 of input image data.Sending part 57 is network interface card or modulator-demodulator etc., and input unit 62 is lithographic plate scanner, Film scanner or digital camera etc.Input unit 62 is reading images optically, generates view data, and the view data that is generated is sent to image compressing device 5, and acceptance division 58 receives the view data of sending from input unit 62.In addition, sending part 57 can send data by communication meanss such as fax or Emails to the outside via not shown communication network.
CPU51 downloads to computer program 61 of the present invention among the RAM52, and the computer program 61 according to downloading carries out the processing of method for compressing image of the present invention.Promptly, at acceptance division 58 under the situation of input unit 62 input image datas, CPU51 carry out with execution mode 1 in foreground layer generating unit 31, reversible compression unit 32, concentration difference dwindle the identical foreground layer of processing that portion 34, non-reversible compression unit 35 and compressed file generating unit 33 carry out and generate processings, reversible compression processing, concentration difference and dwindle processing, non-reversible compression processing and compressed file generation processing.Thus, CPU51 generates the processing of the compressed file that the view data that receives is compressed.CPU51 is stored in the compressed file that is generated in the storage part 54.In addition, CPU51 is according to the computer program of downloading 61, makes sending part 57 send the compressed file that generated or the processing of the compressed file read from storage part 54 to the outside.
As mentioned above, in the present embodiment, also identical with execution mode 1 and 2, when compression comprises the original image of black literal, the edge that detects black literal is as foreground layer, the edge by dwindling the black literal in the original image and the concentration difference of other parts, layer as a setting can improve the visuality of the black literal in the image of compressed file of stretching, extension.Thereby the image compressing device of present embodiment can generate the compressed file of the visuality of literal when improving reproduced image when reducing the size of compressed file.
In addition, the recording medium of the present invention 6 that writes down computer program 61 of the present invention can be any form of CDs such as tape, disk, removable hard disk, CD-ROM/MO/MD/DVD or IC-card (comprising storage card)/card type recording mediums such as light-card.In addition, recording medium 6 of the present invention can be that dressing is in image compressing device 5, CPU51 can reading and recording medium 6 the semiconductor memory of recorded content, that is, mask rom, EPROM (Erasable Programmable Read Only Memory EPROM), EEPROM (Electrically Erasable Read Only Memory), flash ROM etc.
In addition, computer program 61 of the present invention can also be via communication networks such as the Internet or LAN, downloads to image compressing device 5 from the not shown external service apparatus that is connected with image compressing device 5, is stored in the form in the storage part 54.Under the situation of this form, be used for downloading computer program 61 necessary programs and be stored in storage part 54 in advance, perhaps, use drive portion 53 to read, be stored in the storage part 54, download to RAM52 as required from the medium of regulation.
In addition, in above execution mode 1~3, represented as the pixel that is included in the foreground layer, extract form with the processing of the corresponding pixel in edge of black literal, but be not limited to this form, the present invention also can be as being included in pixel in the foreground layer, and extracting with color of pixel is the form of processing of the corresponding pixel of lines of black.In addition, the present invention also can be as being included in pixel in the foreground layer, extracts the form with the processing of black literal and the black corresponding pixel of lines.
This explanation can be implemented under the situation of the spirit that does not break away from fundamental characteristics in a variety of forms particularly, and therefore, above execution mode just is used for explanation rather than is limited to these modes.This is that therefore, all variations that belong to or be equal in the claim scope all are included in the claim because scope of the present invention defines by additional claim definition rather than by specification of the present invention.
Claims (8)
1. image compressing device, this image compressing device compresses the original image that is made of a plurality of pixels, it is characterized in that comprising:
The 1st image production part, detecting color of pixel from original image is the edge of the literal and/or the stick figure of black, generates the 1st image of representing detected edge with bianry image;
Reversible compression unit carries out reversible compression to the 1st image that generates;
The 2nd image production part, generate the 2nd image, the 2nd image is so that in the pixel that original image comprised, the mode that the concentration difference of color diminishes between the pixel that is comprised by the detected edge of described the 1st image production part and other pixel is adjusted and the image that obtains the color of the picture element that original image comprised;
Non-reversible compression unit carries out non-reversible compression to the 2nd image that generates; With
The compressed file generating unit, generation will be through the 1st image of reversible compression and the compressed file that gathers through the 2nd image of non-reversible compression.
2. image compressing device according to claim 1 is characterized in that:
Described the 2nd image production part is included in the concentration value of the concentration of the color of pixel in the detected edge of described the 1st image production part with expression, replace with eight neighborhoods that are positioned at this pixel pixel around concentration value average of pixel.
3. image compressing device according to claim 1 and 2 is characterized in that:
Described the 2nd image pattern looks like into portion the concentration value that is not included in the pixel in the detected edge of described the 1st image production part is carried out smoothing processing.
4. according to each described image compressing device of claim 1~3, it is characterized in that, also comprise:
Low resolution portion carries out the low resolution processing to carrying out non-reversible compression image before.
5. according to each described image compressing device of claim 1~4, it is characterized in that, also comprise:
The processing of the concentration value of each pixel for the original image before adjusting color of pixel by described the 2nd image production part, is revised by the 1st concentration correction portion, makes the concentration of each color of pixel become higher; With
The processing of the concentration value of each pixel for the 2nd image before the non-reversible compression, is revised by the 2nd concentration correction portion, makes the concentration of each color of pixel become higher.
6. according to each described image compressing device of claim 1~5, it is characterized in that:
Described the 1st image production part is included in the color depth of the image in the detected edge to expression the mean value of concentration value calculates,
Also comprise:
The color determination section will be the described mean value that is calculated through the color assignment of the 1st image of reversible compression.
7. compressed image output device is characterized in that comprising:
The described image compressing device of each of claim 1~6;
The image acceptance division, the original image that reception should be compressed by this image compressing device; With
Efferent will export the outside to by the compressed file that described image compressing device generates.
8. method for compressing image, this method for compressing image compresses the original image that is made of a plurality of pixels, it is characterized in that comprising:
Detect the step at the edge of literal that color of pixel is a black and/or stick figure from original image;
The step of the 1st image at detected edge is represented in generation with bianry image;
The step that the 1st image that generates is carried out reversible compression;
Generate the step of the 2nd image, the 2nd image is so that in the pixel that original image comprised, the mode that the concentration difference of color diminishes between the pixel that detected edge comprised and other pixel is adjusted and the image that obtains the color of the picture element that original image comprised;
Generate to adjust the 2nd image that is included in the color of pixel in the original image, make in the pixel that original image comprises, be included in the step that the concentration difference of color between pixel in the detected edge and other pixel is dwindled;
The step that the 2nd image that generates is carried out non-reversible compression; With
The step of the compressed file that generation will gather through the 1st image of reversible compression and through the 2nd image of non-reversible compression.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2010008380A JP5089713B2 (en) | 2010-01-18 | 2010-01-18 | Image compression apparatus, compressed image output apparatus, image compression method, computer program, and recording medium |
JP2010-008380 | 2010-01-18 |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102131038A true CN102131038A (en) | 2011-07-20 |
CN102131038B CN102131038B (en) | 2014-06-25 |
Family
ID=44268913
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201110025301.6A Expired - Fee Related CN102131038B (en) | 2010-01-18 | 2011-01-18 | Image compressing apparatus, compressed image outputting apparatus and image compression method |
Country Status (3)
Country | Link |
---|---|
US (1) | US8463032B2 (en) |
JP (1) | JP5089713B2 (en) |
CN (1) | CN102131038B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105335345A (en) * | 2014-07-03 | 2016-02-17 | 北大方正集团有限公司 | Color image file synthesis method and apparatus |
CN110515658A (en) * | 2014-07-30 | 2019-11-29 | 莫维迪厄斯有限公司 | Method and apparatus for managing variable length instruction |
US11579872B2 (en) | 2013-08-08 | 2023-02-14 | Movidius Limited | Variable-length instruction buffer management |
US11768689B2 (en) | 2013-08-08 | 2023-09-26 | Movidius Limited | Apparatus, systems, and methods for low power computational imaging |
Families Citing this family (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
RU2653461C2 (en) * | 2014-01-21 | 2018-05-08 | Общество с ограниченной ответственностью "Аби Девелопмент" | Glare detection in the image data frame |
JP5807489B2 (en) * | 2011-09-28 | 2015-11-10 | カシオ計算機株式会社 | Image processing apparatus, image processing method, and program |
AU2012201684A1 (en) * | 2012-03-21 | 2013-10-10 | Canon Kabushiki Kaisha | Image compression |
JP5443533B2 (en) * | 2012-03-22 | 2014-03-19 | 株式会社東芝 | Image processing apparatus, image display apparatus, and image processing method |
JP2014078808A (en) | 2012-10-09 | 2014-05-01 | Toshiba Corp | Image processing apparatus and image processing method |
JP5847063B2 (en) * | 2012-11-27 | 2016-01-20 | 京セラドキュメントソリューションズ株式会社 | Image processing device |
JP2015158728A (en) * | 2014-02-21 | 2015-09-03 | 東芝テック株式会社 | Apparatus and program for browsing information |
JP5990217B2 (en) * | 2014-05-21 | 2016-09-07 | キヤノン株式会社 | Image processing apparatus, image forming apparatus, and control method thereof |
US10015504B2 (en) * | 2016-07-27 | 2018-07-03 | Qualcomm Incorporated | Compressing image segmentation data using video coding |
CN106599818B (en) * | 2016-12-07 | 2020-10-27 | 广州视源电子科技股份有限公司 | Method and device for generating handwriting format file based on picture |
US10341671B2 (en) * | 2017-04-21 | 2019-07-02 | Xerox Corporation | Method and system for image compression |
CN113365071B (en) * | 2021-06-08 | 2023-01-17 | 北京数科网维技术有限责任公司 | Image layered compression method and image layered compression device |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH05328143A (en) * | 1992-03-16 | 1993-12-10 | Tokyo Electric Co Ltd | Color piclture coding device |
JP2000354170A (en) * | 1999-06-11 | 2000-12-19 | Seiko Epson Corp | Image compressor, its method and printer with the same image compressor |
JP2005204207A (en) * | 2004-01-19 | 2005-07-28 | Ricoh Co Ltd | Image processing apparatus, image processing program, and storage medium |
JP2005210570A (en) * | 2004-01-26 | 2005-08-04 | Ricoh Co Ltd | Image processing device, image processing program, and memory medium |
CN1747525A (en) * | 2004-09-08 | 2006-03-15 | 佳能株式会社 | Image encoding apparatus and method |
Family Cites Families (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7130442B2 (en) * | 1996-10-28 | 2006-10-31 | International Business Machines Corporation | Protecting images with an image watermark |
EP1435519B1 (en) * | 1997-04-07 | 2007-02-14 | Fisher BioImage ApS | A method for screening substances for effect on cAMP levels based on intracellular translocation of PKA |
US6735341B1 (en) * | 1998-06-18 | 2004-05-11 | Minolta Co., Ltd. | Image processing device and method and recording medium for recording image processing program for same |
JP3558893B2 (en) * | 1998-10-08 | 2004-08-25 | シャープ株式会社 | Image processing apparatus and image processing method |
EP1119186A3 (en) * | 2000-01-19 | 2002-07-31 | Xerox Corporation | Method for generating anti-aliased text and line graphics in compressed document images |
JP4693289B2 (en) | 2001-07-03 | 2011-06-01 | キヤノン株式会社 | Image compression apparatus, image compression method, program code, and storage medium |
US7133565B2 (en) * | 2000-08-25 | 2006-11-07 | Canon Kabushiki Kaisha | Image processing apparatus and method |
US6741755B1 (en) * | 2000-12-22 | 2004-05-25 | Microsoft Corporation | System and method providing mixture-based determination of opacity |
US20040165081A1 (en) * | 2002-12-05 | 2004-08-26 | Hiroyuki Shibaki | Image processing apparatus, image processing system, and image processing method |
JP2004228811A (en) * | 2003-01-21 | 2004-08-12 | Ricoh Co Ltd | Image processor, image processing method and program for computer to perform execution |
US7295700B2 (en) * | 2003-10-24 | 2007-11-13 | Adobe Systems Incorporated | Object extraction based on color and visual texture |
JP2005204206A (en) * | 2004-01-19 | 2005-07-28 | Ricoh Co Ltd | Image processing apparatus, image processing program, and storage medium |
JP2005259017A (en) * | 2004-03-15 | 2005-09-22 | Ricoh Co Ltd | Image processing apparatus, image processing program and storage medium |
US7873218B2 (en) * | 2004-04-26 | 2011-01-18 | Canon Kabushiki Kaisha | Function approximation processing method and image processing method |
US7450268B2 (en) * | 2004-07-02 | 2008-11-11 | Hewlett-Packard Development Company, L.P. | Image reproduction |
JP4646797B2 (en) * | 2005-02-01 | 2011-03-09 | キヤノン株式会社 | Image processing apparatus, control method therefor, and program |
US7783117B2 (en) * | 2005-08-12 | 2010-08-24 | Seiko Epson Corporation | Systems and methods for generating background and foreground images for document compression |
JP2007089136A (en) * | 2005-08-23 | 2007-04-05 | Ricoh Co Ltd | Image processing method, image processing program, recording medium, and composite device |
JP5328143B2 (en) * | 2007-06-18 | 2013-10-30 | ジーオーピー株式会社 | Portable worktable |
US20090046938A1 (en) * | 2007-08-15 | 2009-02-19 | Lockheed Martin Corporation | Character contour correction |
-
2010
- 2010-01-18 JP JP2010008380A patent/JP5089713B2/en active Active
-
2011
- 2011-01-18 US US13/008,378 patent/US8463032B2/en not_active Expired - Fee Related
- 2011-01-18 CN CN201110025301.6A patent/CN102131038B/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH05328143A (en) * | 1992-03-16 | 1993-12-10 | Tokyo Electric Co Ltd | Color piclture coding device |
JP2000354170A (en) * | 1999-06-11 | 2000-12-19 | Seiko Epson Corp | Image compressor, its method and printer with the same image compressor |
JP2005204207A (en) * | 2004-01-19 | 2005-07-28 | Ricoh Co Ltd | Image processing apparatus, image processing program, and storage medium |
JP2005210570A (en) * | 2004-01-26 | 2005-08-04 | Ricoh Co Ltd | Image processing device, image processing program, and memory medium |
CN1747525A (en) * | 2004-09-08 | 2006-03-15 | 佳能株式会社 | Image encoding apparatus and method |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11579872B2 (en) | 2013-08-08 | 2023-02-14 | Movidius Limited | Variable-length instruction buffer management |
US11768689B2 (en) | 2013-08-08 | 2023-09-26 | Movidius Limited | Apparatus, systems, and methods for low power computational imaging |
CN105335345A (en) * | 2014-07-03 | 2016-02-17 | 北大方正集团有限公司 | Color image file synthesis method and apparatus |
CN105335345B (en) * | 2014-07-03 | 2017-12-19 | 北大方正集团有限公司 | Color document image synthetic method and device |
CN110515658A (en) * | 2014-07-30 | 2019-11-29 | 莫维迪厄斯有限公司 | Method and apparatus for managing variable length instruction |
Also Published As
Publication number | Publication date |
---|---|
US20110176728A1 (en) | 2011-07-21 |
CN102131038B (en) | 2014-06-25 |
JP2011147085A (en) | 2011-07-28 |
JP5089713B2 (en) | 2012-12-05 |
US8463032B2 (en) | 2013-06-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102131038B (en) | Image compressing apparatus, compressed image outputting apparatus and image compression method | |
CN101902549B (en) | Image processing apparatus and image processing method | |
CN101901487B (en) | Image compressing method, image compressing apparatus, and image forming apparatus | |
CN102223464B (en) | Image compressing apparatus, image outputting apparatus, and image compressing method | |
CN101616231B (en) | Image processing apparatus, image forming apparatus and image processing method | |
CN101938650B (en) | Image compressing apparatus, image compressing method, image decompressing apparatus, image decompressing method, image forming apparatus | |
CN101640754B (en) | Image compressing method, image compressing apparatus and image forming apparatus | |
CN101790023B (en) | Image compressing method, image compressing apparatus and image forming apparatus | |
JP4528846B2 (en) | Image compression method, image compression apparatus, image processing apparatus, image forming apparatus, computer program, and recording medium | |
CN104243768B (en) | Image processing apparatus and image processing system | |
JP2010045537A (en) | Image processor, image forming device, image processing method, computer program and recording medium | |
JP5259862B2 (en) | Image processing apparatus, image forming apparatus including the same, program, and recording medium | |
JP4926118B2 (en) | Image compression apparatus, image processing apparatus, image forming apparatus, computer program, and recording medium | |
JP2012231419A (en) | Image compression apparatus, image forming apparatus, image reading device, image compression method, computer program, and recording medium | |
JP2010187179A (en) | Image processing apparatus, image forming apparatus, image processing method, computer program, and recording medium | |
JP2004214908A (en) | Image processing apparatus and image forming apparatus provided with the same | |
JP2009017317A (en) | Image processing method, image processing device, image forming apparatus, image processing program and recording medium with image processing program recorded thereon | |
JP2010098611A (en) | Image processing apparatus, image forming apparatus, image processing method, computer program, and recording medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140625 Termination date: 20210118 |
|
CF01 | Termination of patent right due to non-payment of annual fee |